SPECIAL MORPHOLOGY : PERNS. 197 



the latter are in a more or less closed circle, be distinguished into 

 medulla in the interior and cortex in the exterior. The vascular 

 bundles lie in their vertical course alternately side by side, so as 

 to form a net, the meshes of which furnish, at their upper part, 

 branches of the bundles to the leaves and branches, where the latter 

 are present ; we find a few isolated vascular bundles in the medulla 

 of the arborescent Ferns, which pass out through those meshes 

 into the leaves. The vascular bundles have often a band-like, or 

 channel-shaped form, compressed from within outwards; and are 

 generally surrounded by a sheath of thick-walled elongated cells 

 coloured brown (by tannin and humic acid ?) ; and bundles of such 

 cells alone also traverse the stem. The parenchymatous cellular 

 walls, on dying, speedily acquire a more or less dark brown colour. 

 It is well known that many Ferns contain a large quantity of tannin. 

 The parenchyma, especially the base of the leaf-stalk, often con- 

 tains much starch (as, for instance, in the Marattia cicutafolia), 

 which serves in some of the South Sea Islands as an article of 

 food. The vascular bundles are composed of porous vessels, having 

 the pores small, or most frequently with slits ; sometimes, however, 

 as in the leaf-stalks, we find spiral vessels which admit of being 

 unrolled. The leaves but rarely consist of one single cellular 

 layer (such being only the case in the HymenophyllecB), but generally 

 of several, forming two laminae : an upper one, composed of short 

 cylindrical cells, vertical to the surface of the leaf, and an under 

 one, formed of loose spherical, or sponge-like, parenchyma. More- 

 over, the two sides are invested with a true epidermis, which always 

 exhibits perfect stomata on its lower surface. The upper epider- 

 mis not unfrequently consists of several layers of cells. Isolated 

 bundles of liber cells are often met with above and below the 

 vascular bundles of the leaves. The leaves contain a large quantity 

 of potash salts. 



The attempt to represent the stem of the Fern as merely composed of 

 leaf-stalks grown together is so entirely at variance with the law 

 of its development, and, consequently, so totally devoid of foundation, that 

 we do not deem it worth while to contest the point. Germination shows 

 that there is a rudiment of the stem prior to the formation of the leaves 

 and leaf-stalks. We may refer for the anatomy of the stem to Mold's 

 work, which we have already mentioned, and which certainly still leaves 

 much to be done, although he can scarcely be blamed for the deficiency. 

 We now feel the want of the history of the living development, and a 

 much greater service would be rendered to science if one of the many 

 travellers in Brazil would furnish us with the result of an accurate inves- 

 tigation of this development in the stem of an arborescent Fern, instead of 

 adding a couple of thousand dried new species to the 80,000 which we 

 already possess, and which are scarcely worthy of notice while we continue, 

 as at present, hardly to have a certain knowledge of any one single speci- 

 men. The annulus of the sporocarp exhibits a structure similar to the 

 teeth of the fruit of the Mosses. I think I have, detected very delicate 

 spiral fibres in the cells of the walls of the fruit in Ophioylossunt 

 and Osmunda. 



o 3 



